{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## NN Hidden Layer\n",
    "\n",
    "Here e will cover\n",
    "- Hidden layers\n",
    "- Non-linear transfer functions\n",
    "- Backprop\n",
    "\n",
    "While the previous tutorials described very simple single layer regression and classification models, this tutorial will describe a 2-class classification neural network with 1 input dimension, and a non-linear hidden layer with 1 neuron.\n",
    "\n",
    "The notebook starts out with importing the libraries we need:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np # Matrix and vector computation package\n",
    "import matplotlib.pyplot as plt  # Plotting library\n",
    "from matplotlib.colors import colorConverter, ListedColormap # some plotting functions\n",
    "from mpl_toolkits.mplot3d import Axes3D  # 3D plots\n",
    "from matplotlib import cm # Colormaps\n",
    "# Allow matplotlib to plot inside this notebook\n",
    "%matplotlib inline\n",
    "# Set the seed of the numpy random number generator so that the tutorial is reproducable\n",
    "np.random.seed(seed=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Define dataset\n",
    "\n",
    "In this example the target classes $t$ corresponding to the inputs $x$ will be generated from 2 class distributions: blue ($t=1$) and red ($t=0$). Where the red class is a [multimodal distribution](http://en.wikipedia.org/wiki/Multimodal_distribution) that surrounds the distribution of the blue class. This results in a 1D dataset that is not linearly separable. These samples are plotted in the figure below.\n",
    "\n",
    "The model from part 2 won't be able to classify both classes correctly since it can learn only linear separators. By adding a hidden layer with a non-linear transfer function, the model will be able to train a non-linear classifier."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define and generate the samples\n",
    "nb_of_samples_per_class = 20  # The number of sample in each class\n",
    "blue_mean = [0]  # The mean of the blue class\n",
    "red_left_mean = [-2]  # The mean of the red class\n",
    "red_right_mean = [2]  # The mean of the red class\n",
    "\n",
    "std_dev = 0.5  # standard deviation of both classes\n",
    "# Generate samples from both classes\n",
    "x_blue = np.random.randn(nb_of_samples_per_class, 1) * std_dev + blue_mean\n",
    "x_red_left = np.random.randn(nb_of_samples_per_class//2, 1) * std_dev + red_left_mean\n",
    "x_red_right = np.random.randn(nb_of_samples_per_class//2, 1) * std_dev + red_right_mean\n",
    "\n",
    "# Merge samples in set of input variables x, and corresponding set of\n",
    "# output variables t\n",
    "x = np.vstack((x_blue, x_red_left, x_red_right))\n",
    "t = np.vstack((np.ones((x_blue.shape[0],1)), \n",
    "               np.zeros((x_red_left.shape[0],1)), \n",
    "               np.zeros((x_red_right.shape[0], 1))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdcAAABdCAYAAAD32CzLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1134ccc18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot samples from both classes as lines on a 1D space\n",
    "plt.figure(figsize=(8,0.5))\n",
    "plt.xlim(-3,3)\n",
    "plt.ylim(-1,1)\n",
    "# Plot samples\n",
    "plt.plot(x_blue, np.zeros_like(x_blue), 'b|', ms = 30) \n",
    "plt.plot(x_red_left, np.zeros_like(x_red_left), 'r|', ms = 30) \n",
    "plt.plot(x_red_right, np.zeros_like(x_red_right), 'r|', ms = 30) \n",
    "plt.gca().axes.get_yaxis().set_visible(False)\n",
    "plt.title('Input samples from the blue and red class')\n",
    "plt.xlabel('$x$', fontsize=15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Non-linear transfer function\n",
    "\n",
    "The non-linear transfer function used in the hidden layer of this example is the [Gaussian radial basis](http://en.wikipedia.org/wiki/Radial_basis_function) function (RBF).\n",
    "\n",
    "The RBF is a transfer function that is not usually used in neural networks, except for [radial basis function networks](http://en.wikipedia.org/wiki/Radial_basis_function_network). One of the most common transfer functions in neural networks is the [sigmoid transfer function](http://en.wikipedia.org/wiki/Sigmoid_function). \n",
    "\n",
    "The RBF will allow to separate the blue samples from the red samples in this simple example by only activating for a certain region around the origin. The RBF is plotted in the figure below and is defined in this example as:\n",
    "\n",
    "$$ \\text{RBF} = \\phi(z) = e^{-z^2} $$\n",
    "\n",
    "The derivative of this RBF function is:\n",
    "\n",
    "$$ \\frac{d \\phi(z)}{dz} = -2 z e^{-z^2} = -2 z \\phi(z)$$ "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Define the rbf function\n",
    "def rbf(z):\n",
    "    return np.exp(-z**2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZAAAAEbCAYAAAD0yNLXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRD4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113449470>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the rbf function\n",
    "z = np.linspace(-6,6,100)\n",
    "plt.plot(z, rbf(z), 'b-')\n",
    "plt.xlabel('$z$', fontsize=15)\n",
    "plt.ylabel('$e^{-z^2}$', fontsize=15)\n",
    "plt.title('RBF function')\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Optimization by backpropagation\n",
    "\n",
    "We will train this model by using the backpropagation algorithm that is typically used to train neural networks. Each iteration of the backpropagation algorithm consists of two steps:\n",
    "A forward propagation step to compute the output of the network.\n",
    "A backward propagation step in which the error at the end of the network is propagated backward through all the neurons while updating their parameters.\n",
    "\n",
    "#### 1. Forward step\n",
    "During the forward step, the input will be propagated layer by layer through the network to compute the final output of the network.\n",
    "\n",
    "##### Compute activations of hidden layer\n",
    "\n",
    "The activations $\\mathbf{h}$ of the hidden layer will be computed by:\n",
    "\n",
    "$$\\mathbf{h} = \\phi(\\mathbf{x}*w_h) = e^{-(\\mathbf{x}*w_h)^2} $$\n",
    "\n",
    "With $w_h$ the weight parameter that transforms the input before applying the RBF transfer function. This is implemented below by the hidden_activations(x, wh) method.\n",
    "\n",
    "##### Compute activations of output\n",
    "\n",
    "The output of the final layer and network will be computed by passing the hidden activations $\\mathbf{h}$ as input to the logistic output function:\n",
    "\n",
    "$$ \\mathbf{y} = \\sigma(\\mathbf{h} * w_o - 1) = \\frac{1}{1+e^{-\\mathbf{h} * w_o - 1}} $$\n",
    "\n",
    "With $w_o$ the weight parameter of the output layer. This is implemented below as the output_activations(h , wo) method. \n",
    "\n",
    "Note that we add a bias (intercept) term of $-1$ to the input of the logistic output neuron. Remember from part 2 that the logistic output neuron without bias can only learn a decision boundary that goes through the origin $(0)$. \n",
    "\n",
    "Since the RBF in the hidden layer projects all input variables to a range between $0$ and $1$, the output layer without an intercept will not be able to learn any useful classifier, because none of the samples will be below $0$ and thus lie on the left side of the decision boundary. By adding a bias term the decision boundary is moved from the intercept. Normally the value of this bias term is learned together with the rest of the weight parameters, but to keep this model simple we just make this bias constant in this example. [(More details on this here.)](http://stackoverflow.com/a/42436987/919431) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the logistic function\n",
    "def logistic(z): \n",
    "    return 1 / (1 + np.exp(-z))\n",
    "\n",
    "# Function to compute the hidden activations\n",
    "def hidden_activations(x, wh):\n",
    "    return rbf(x * wh)\n",
    "\n",
    "# Define output layer feedforward\n",
    "def output_activations(h , wo):\n",
    "    return logistic(h * wo - 1)\n",
    "\n",
    "# Define the neural network function\n",
    "def nn(x, wh, wo): \n",
    "    return output_activations(hidden_activations(x, wh), wo)\n",
    "\n",
    "# Define the neural network prediction function that only returns\n",
    "#  1 or 0 depending on the predicted class\n",
    "def nn_predict(x, wh, wo): \n",
    "    return np.around(nn(x, wh, wo))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  2. Backward step\n",
    "The backward step will begin with computing the cost at the output node. This cost will then be propagated backwards layer by layer through the network to update the parameters.\n",
    "\n",
    "The gradient descent algorithm is used in every layer to update the parameters in the direction of the negative gradient.\n",
    "\n",
    "The parameters $w_h$ and $w_o$ are updated by $w(k+1) = w(k) - \\Delta w(k+1)$. $\\Delta w$ is defined as: $\\Delta w = \\mu * {\\partial \\xi}/{\\partial w}$ with $\\mu$ the learning rate and ${\\partial \\xi}/{\\partial w}$ the gradient of the parameter $w$ with respect to the cost function $\\xi$.\n",
    "\n",
    "##### Compute the cost function\n",
    "\n",
    "The cost function $\\xi$ used in this model is the same cross-entropy cost function explained in Logistic & Softmax Tutorial:\n",
    "\n",
    "$$\\xi(t_i,y_i) = - \\left[ t_i log(y_i) + (1-t_i)log(1-y_i) \\right]$$\n",
    "\n",
    "This cost function is plotted for the $w_h$ and $w_o$ parameters in the next figure. Note that this error surface is not convex anymore and that the $w_h$ parameter mirrors the cost function along the $w_h = 0$ axis.\n",
    "\n",
    "Also, notice that this cost function has a very sharp gradient around $w_h = 0$ starting from $w_o>0$ and that the minima run along the lower edge of this peak. If the learning rate will be to big, the updates might jump over the minima gap, onto the sharp gradient. Because the gradient is sharp, the update will be large, and we might end up further from the minima than we started."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Define the cost function\n",
    "def cost(y, t):\n",
    "    return - np.sum(np.multiply(t, np.log(y)) + np.multiply((1-t), np.log(1-y)))\n",
    "\n",
    "# Define a function to calculate the cost for a given set of parameters\n",
    "def cost_for_param(x, wh, wo, t):\n",
    "    return cost(nn(x, wh, wo) , t)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcAAAAE1CAYAAABwejDZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11394f048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the cost in function of the weights\n",
    "# Define a vector of weights for which we want to plot the cost\n",
    "nb_of_ws = 200 # compute the cost nb_of_ws times in each dimension\n",
    "wsh = np.linspace(-10, 10, num=nb_of_ws) # hidden weights\n",
    "wso = np.linspace(-10, 10, num=nb_of_ws) # output weights\n",
    "ws_x, ws_y = np.meshgrid(wsh, wso) # generate grid\n",
    "cost_ws = np.zeros((nb_of_ws, nb_of_ws)) # initialize cost matrix\n",
    "# Fill the cost matrix for each combination of weights\n",
    "for i in range(nb_of_ws):\n",
    "    for j in range(nb_of_ws):\n",
    "        cost_ws[i,j] = cost(nn(x, ws_x[i,j], ws_y[i,j]) , t)\n",
    "# Plot the cost function surface\n",
    "fig = plt.figure()\n",
    "ax = Axes3D(fig)\n",
    "# plot the surface\n",
    "surf = ax.plot_surface(ws_x, ws_y, cost_ws, linewidth=0, cmap=cm.pink)\n",
    "ax.view_init(elev=60, azim=-30)\n",
    "cbar = fig.colorbar(surf)\n",
    "ax.set_xlabel('$w_h$', fontsize=15)\n",
    "ax.set_ylabel('$w_o$', fontsize=15)\n",
    "ax.set_zlabel('$\\\\xi$', fontsize=15)\n",
    "cbar.ax.set_ylabel('$\\\\xi$', fontsize=15)\n",
    "plt.title('Cost function surface')\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Update the output layer\n",
    "\n",
    "At the output the gradient for sample $i$, ${\\partial \\xi_i}/{\\partial w_o}$, can be worked out the same way as we did before:\n",
    "\n",
    "$$\\frac{\\partial \\xi_i}{\\partial w_o} = \\frac{\\partial z_{oi}}{\\partial w_o} \\frac{\\partial y_i}{\\partial z_{oi}} \\frac{\\partial \\xi_i}{\\partial y_i} = h_i (y_i-t_i) = h_i * \\delta_{oi}$$\n",
    "\n",
    "With $z_{oi} = h_i * w_o$, $h_i$ the hidden layer activation of sample $i$ and ${\\partial \\xi_i}/{\\partial z_{oi}} = \\delta_{oi}$ the gradient of the error at the output layer of the neural network with respect to the input to this layer.\n",
    "\n",
    "$\\delta_{o}$ is defined below as the gradient_output(y, t) method and ${\\partial \\xi}/{\\partial w_o}$ as the gradient_weight_out(h, grad_output) method.\n",
    "\n",
    "##### Update the hidden layer\n",
    "\n",
    "At the hidden layer the gradient for sample $i$, ${\\partial \\xi_i}/{\\partial w_{h}}$, of the hidden neuron is computed the same way:\n",
    "\n",
    "$$\\frac{\\partial \\xi_i}{\\partial w_{h}} = \\frac{\\partial z_{hi}}{\\partial w_{h}} \\frac{\\partial h_i}{\\partial z_{hi}} \\frac{\\partial \\xi_i}{\\partial h_i}$$\n",
    "\n",
    "With $z_{hi} = x_i * w_{h} $. And with ${\\partial \\xi_i}/{\\partial z_{hi}} = \\delta_{hi}$ the gradient of the error at the input of the hidden layer with respect to the input to this layer. This error can be interpreted as the contribution of $z_{hi}$ to the final error. How do we define this error gradient $\\delta_{hi}$ at the input of the hidden neurons? It can be computed as the error gradient propagated back from the output layer through the hidden layer.\n",
    "\n",
    "$$\\delta_{hi} = \\frac{\\partial \\xi_i}{\\partial z_{hi}} = \\frac{\\partial h_i}{\\partial z_{hi}} \\frac{\\partial z_{oi}}{\\partial h_i} \\frac{\\partial \\xi_i}{\\partial z_{oi}} \n",
    "= (-2 * z_{hi} * h_i) * w_{o} * (y_i - t_i) = -2 * z_{hi} * h_i * w_{o} * \\delta_{oi} $$\n",
    "\n",
    "Because of this, and because ${\\partial z_{hi}}/{\\partial w_{h}} = x_i$ we can compute ${\\partial \\xi_i}/{\\partial w_{h}}$ as:\n",
    "\n",
    "$$\\frac{\\partial \\xi_i}{\\partial w_{h}} = x_i \\delta_{hi}  $$\n",
    "\n",
    "The gradients for each parameter can again be summed up to compute the update for a batch of input examples.\n",
    "$\\delta_{h}$ is defined below as the gradient_hidden(wo, grad_output) method and ${\\partial \\xi}/{\\partial w_h}$ as the gradient_weight_hidden(x, zh, h, grad_hidden) method.\n",
    "\n",
    "To start out the gradient descent algorithm, you typically start with picking the initial parameters at random and start updating these parameters in the direction of the negative gradient with help of the backpropagation algorithm. One backpropagation iteration is implemented below by the backprop_update(x, t, wh, wo, learning_rate) method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Define the error function\n",
    "def gradient_output(y, t):\n",
    "    return y - t\n",
    "\n",
    "# Define the gradient function for the weight parameter at the output layer\n",
    "def gradient_weight_out(h, grad_output): \n",
    "    return  h * grad_output\n",
    "\n",
    "# Define the gradient function for the hidden layer\n",
    "def gradient_hidden(wo, grad_output):\n",
    "    return wo * grad_output\n",
    "\n",
    "# Define the gradient function for the weight parameter at the hidden layer\n",
    "def gradient_weight_hidden(x, zh, h, grad_hidden):\n",
    "    return x * -2 * zh * h * grad_hidden\n",
    "\n",
    "# Define the update function to update the network parameters over 1 iteration\n",
    "def backprop_update(x, t, wh, wo, learning_rate):\n",
    "    # Compute the output of the network\n",
    "    # This can be done with y = nn(x, wh, wo), but we need the intermediate \n",
    "    #  h and zh for the weight updates.\n",
    "    zh = x * wh\n",
    "    h = rbf(zh)  # hidden_activations(x, wh)\n",
    "    y = output_activations(h, wo)\n",
    "    # Compute the gradient at the output\n",
    "    grad_output = gradient_output(y, t)\n",
    "    # Get the delta for wo\n",
    "    d_wo = learning_rate * gradient_weight_out(h, grad_output)\n",
    "    # Compute the gradient at the hidden layer\n",
    "    grad_hidden = gradient_hidden(wo, grad_output)\n",
    "    # Get the delta for wh\n",
    "    d_wh = learning_rate * gradient_weight_hidden(x, zh, h, grad_hidden)\n",
    "    # return the update parameters\n",
    "    return (wh-d_wh.sum(), wo-d_wo.sum())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Backprop updates\n",
    "\n",
    "An example run of backpropagation for 50 iterations on the example inputs $\\mathbf{x}$ and targets $\\mathbf{t}$ is shown in the figure below. The white dots represent the weight parameter values $w_h$ and $w_o$ at iteration $k$ and are plotted on the cost surface.\n",
    "\n",
    "Notice that we decrease the learning rate linearly with each step. This is to make sure that in the end the learning rate is 0 and the sharp gradient will not allow the weight paramaters to fluctuate much during the last few iterations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "final cost is 9.16 for weights wh: 1.44 and wo: 5.77\n"
     ]
    }
   ],
   "source": [
    "# Run backpropagation\n",
    "# Set the initial weight parameter\n",
    "wh = 2\n",
    "wo = -5\n",
    "# Set the learning rate\n",
    "learning_rate = 0.2\n",
    "\n",
    "# Start the gradient descent updates and plot the iterations\n",
    "nb_of_iterations = 50  # number of gradient descent updates\n",
    "lr_update = learning_rate / nb_of_iterations # learning rate update rule\n",
    "w_cost_iter = [(wh, wo, cost_for_param(x, wh, wo, t))]  # List to store the weight values over the iterations\n",
    "for i in range(nb_of_iterations):\n",
    "    learning_rate -= lr_update # decrease the learning rate\n",
    "    # Update the weights via backpropagation\n",
    "    wh, wo = backprop_update(x, t, wh, wo, learning_rate) \n",
    "    w_cost_iter.append((wh, wo, cost_for_param(x, wh, wo, t)))  # Store the values for plotting\n",
    "\n",
    "# Print the final cost\n",
    "print('final cost is {:.2f} for weights wh: {:.2f} and wo: {:.2f}'.format(cost_for_param(x, wh, wo, t), wh, wo))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcAAAAE1CAYAAABwejDZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1138aa908>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the weight updates on the error surface\n",
    "# Plot the error surface\n",
    "fig = plt.figure()\n",
    "ax = Axes3D(fig)\n",
    "surf = ax.plot_surface(ws_x, ws_y, cost_ws, linewidth=0, cmap=cm.pink)\n",
    "ax.view_init(elev=60, azim=-30)\n",
    "cbar = fig.colorbar(surf)\n",
    "cbar.ax.set_ylabel('$\\\\xi$', fontsize=15)\n",
    "\n",
    "# Plot the updates\n",
    "for i in range(1, len(w_cost_iter)):\n",
    "    wh1, wo1, c1 = w_cost_iter[i-1]\n",
    "    wh2, wo2, c2 = w_cost_iter[i]\n",
    "    # Plot the weight-cost value and the line that represents the update \n",
    "    ax.plot([wh1], [wo1], [c1], 'w+')  # Plot the weight cost value\n",
    "    ax.plot([wh1, wh2], [wo1, wo2], [c1, c2], 'w-')\n",
    "# Plot the last weights\n",
    "wh1, wo1, c1 = w_cost_iter[len(w_cost_iter)-1]\n",
    "ax.plot([wh1], [wo1], c1, 'w+')\n",
    "# Shoz figure\n",
    "ax.set_xlabel('$w_h$', fontsize=15)\n",
    "ax.set_ylabel('$w_o$', fontsize=15)\n",
    "ax.set_zlabel('$\\\\xi$', fontsize=15)\n",
    "plt.title('Gradient descent updates on cost surface')\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualizing trained classifier\n",
    "\n",
    "The resulting decision boundary of running backpropagation on the example inputs $\\mathbf{x}$ and targets $\\mathbf{t}$ is shown in the figure below. The background color (blue, red) refers to the classification decision of the trained classifier at that position in the input space. Note that all examples are classified correctly by the trained classifier."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdcAAABYCAYAAACnFb14AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1139499b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the resulting decision boundary\n",
    "# Generate a grid over the input space to plot the color of the\n",
    "#  classification at that grid point\n",
    "nb_of_xs = 100\n",
    "xs = np.linspace(-3, 3, num=nb_of_xs)\n",
    "ys = np.linspace(-1, 1, num=nb_of_xs)\n",
    "xx, yy = np.meshgrid(xs, ys) # create the grid\n",
    "# Initialize and fill the classification plane\n",
    "classification_plane = np.zeros((nb_of_xs, nb_of_xs))\n",
    "for i in range(nb_of_xs):\n",
    "    for j in range(nb_of_xs):\n",
    "        classification_plane[i,j] = nn_predict(xx[i,j], wh, wo)\n",
    "# Create a color map to show the classification colors of each grid point\n",
    "cmap = ListedColormap([\n",
    "        colorConverter.to_rgba('r', alpha=0.25),\n",
    "        colorConverter.to_rgba('c', alpha=0.25)])\n",
    "\n",
    "# Plot the classification plane with decision boundary and input samples\n",
    "plt.figure(figsize=(8,0.5))\n",
    "plt.contourf(xx, yy, classification_plane, cmap=cmap)\n",
    "plt.xlim(-3,3)\n",
    "plt.ylim(-1,1)\n",
    "# Plot samples from both classes as lines on a 1D space\n",
    "plt.plot(x_blue, np.zeros_like(x_blue), 'c|', ms = 30) \n",
    "plt.plot(x_red_left, np.zeros_like(x_red_left), 'r|', ms = 30) \n",
    "plt.plot(x_red_right, np.zeros_like(x_red_right), 'r|', ms = 30) \n",
    "plt.gca().axes.get_yaxis().set_visible(False)\n",
    "plt.title('Input samples and their classification')\n",
    "plt.xlabel('x')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Transformation of the input domain\n",
    "\n",
    "How is the neural network able to separate the non-linearly seperable classes with a linear logistic classifier at the output? The key is the hidden layer with the non-linear RBF transfer function. Note that the RBF transfer function is able to transform the samples near the origin (blue class) to a value larger than $0$, and the samples further from the origin (red samples) to a value near $0$. This projection is plotted in the following figure. Note that the red samples are located around $0$ to the left, and that the blue samples are located more to the right. This projection is linearly seperable by the logistic classifier in the output layer.\n",
    "\n",
    "Also, note that the offset of the peak of the Gaussian function we use is $0$. This means that the Gaussian function is centered around the origin, which can be noted in the symmetrical decision boundaries around the origin on the previous figure. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdgAAABYCAYAAABWHub1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113999ac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot projected samples from both classes as lines on a 1D space\n",
    "plt.figure(figsize=(8,0.5))\n",
    "plt.xlim(-0.01,1)\n",
    "plt.ylim(-1,1)\n",
    "# Plot projected samples\n",
    "plt.plot(hidden_activations(x_blue, wh), np.zeros_like(x_blue), 'c|', ms = 30) \n",
    "plt.plot(hidden_activations(x_red_left, wh), np.zeros_like(x_red_left), 'r|', ms = 30) \n",
    "plt.plot(hidden_activations(x_red_right, wh), np.zeros_like(x_red_right), 'r|', ms = 30) \n",
    "plt.gca().axes.get_yaxis().set_visible(False)\n",
    "plt.title('Projection of the input samples by the hidden layer.')\n",
    "plt.xlabel('h')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}